JPMorgan SRE AVP Interview Experience: 5 Rounds Decoded 2026

SRE AVP at JPMorganChase. Five rounds. A HackerRank session the candidate described as "hell." Ten days of silence after the HR discussion — then the offer. In a LinkedIn post that drew 140 reactions, ex-Verizon, ex-Infosys engineer Praveen Singampalli shared the complete arc, from coding grind to formal offer letter. Here is every detail that matters.

The 5-Round JPMC SRE AVP Process, Broken Down Round by Round

Praveen Singampalli is an AWS Community Builder with prior stints at Verizon and Infosys. He shared his JPMorganChase SRE AVP interview experience in a public LinkedIn post; the five-round structure he describes is the basis for everything below.

Round 1 — HackerRank Coding (Proctored)

No warm-up. JPMC opens the SRE AVP process with a timed, proctored HackerRank assessment. Praveen was direct: "Trust me the coding round was hell." For a role that carries "Site Reliability Engineer" in the title, the expectation is not basic scripting — it is algorithmic depth and timed execution under pressure. Candidates who treat SRE as "Linux plus YAML" and skip DSA preparation routinely drop out here. This round is the primary filter; it gates access to every technical conversation that follows.

Rounds 2 and 3 — Technical Interviews

The source does not surface every question from the mid-rounds, but JPMC SRE interviews at AVP level consistently cover Kubernetes internals, container networking, Linux observability, and distributed-systems failure modes. At AVP level the question frame shifts from "what would you do?" to "walk me through an RCA you personally owned." Depth of ownership is what distinguishes passing answers from strong ones.

Round 4 — Behavioral and Situational (STAR Format)

Praveen's own description of this round: "tricky as we have to respond every answer in STAR format and few situational based and real time challenges were thrown to me." JPMC runs structured behavioral interviews at VP and AVP levels. Every answer must be delivered in Situation → Task → Action → Result format. Improvised stories that skip measurable results do not survive the follow-up probing. Real-time incident scenarios also appeared — the kind of prompt that sounds like "your Kubernetes pod is OOMKilling in prod at 2 AM, walk me through the next thirty minutes."

Round 5 — HR and Offer Discussion

After four rounds, Praveen entered the compensation discussion. The formal offer letter then took 10 days to arrive. JPMC's HR machinery handles thousands of hires across global entities; that window is normal process, not a negative signal.

Round	Format	Primary Focus
1	HackerRank (proctored)	DSA / coding under time pressure
2	Technical interview	Linux, systems, infra depth
3	Technical interview	Kubernetes, networking, system design
4	Behavioral (STAR)	Incident ownership, situational judgment
5	HR discussion	Offer and compensation

The DevOps and SRE Prep Roadmap Praveen Would Use If Starting Over in 2026

Alongside the interview breakdown, Praveen published a structured prep framework on LinkedIn. It is worth treating as a calibration checklist, not a reading list.

Stage 1 — Linux as a Diagnostic Tool, Not a Command Catalogue

The distinction he draws is precise: Linux mastery is not memorising flags, it is reading system state from first principles. The specific commands he cites — ps aux | grep java, netstat -tulpn, dmesg | tail -20 — are post-incident instruments. An engineer who can reconstruct what went wrong from logs alone clears a bar that most candidates miss at the first technical round.

For JPMC SRE specifically: expect questions on process trees, signal handling, file descriptor limits, and /proc entries at AVP level. Describing what these are is insufficient — you need to have used them to diagnose something real.

Stage 2 — Network Tracing, Not Network Trivia

"Every outage will test you on networking," Praveen writes. The framing he uses is specific: can you trace a DNS failure end-to-end? Can you identify where a TCP reset propagates from a single pod to a checkout API? These are Kubernetes networking questions disguised as SRE judgment questions. Build a local lab. Run tcpdump between pods. Break DNS in a k3s cluster and fix it without documentation. Interviewers at AVP level distinguish people who have memorised the OSI model from people who have actually seen packets drop.

Stage 3 — Containers: Internals Over YAML

Docker and Kubernetes at SRE AVP level means understanding cgroups, namespaces, image layer caching, and OOMKill sequencing. kubectl describe pod is table stakes. Knowing why a pod's memory request and limit diverge — and what that means for the node's QoS class and eviction order — is the AVP signal.

Prep Area	Entry-Level Focus	AVP-Level Focus
Linux	Commands and flags	/proc, signal handling, fd limits, log-based RCA
Networking	OSI model, subnets	tcpdump, pod-to-pod DNS failure, TCP reset tracing
Containers	Docker build, kubectl basics	cgroups, namespaces, OOMKill, Kubernetes QoS classes
Behavioral	Generic STAR stories	Named incidents, measurable outcomes, deep follow-up answers

Cracking JPMC's STAR Behavioral Round Without Sounding Scripted

The STAR framework is widely known. The failure mode is equally predictable: answers that are structurally STAR-shaped but feel rehearsed, thin, and free of real detail. JPMC's behavioral interviewers at AVP level probe the "Action" segment hard — expect "why that approach over the alternative?" follow-ups on every story you tell.

Build a story bank before the first screen:

• 3–5 production incidents you personally diagnosed and resolved, with specific tools, specific timelines, and measurable blast-radius numbers
• 2–3 cross-team conflict resolutions involving a concrete constraint — a deadline, a resource contention, a competing product priority
• 2–3 examples of proactive improvement (not reactive firefighting) — monitoring gaps you closed, capacity planning you initiated, pipeline latency you reduced by a named percentage

For real-time incident scenario questions:

Narrate your decision tree out loud. JPMC interviewers are not looking for the single correct answer — they are watching how you structure uncertainty and whether you escalate at the right threshold. Practice by setting a timer, reading a scenario, and speaking your response start-to-finish without pausing to type notes.

Timing:

Target 2 to 2.5 minutes per STAR answer. Under 90 seconds reads as thin. Over 3 minutes loses the room. Praveen's note that the behavioral round was "tricky" likely reflects the combination of strict format enforcement and live curveball scenarios in the same sitting.

Common Mistakes That Kill JPMC SRE Candidacies Early

• Skipping DSA because it is an infrastructure role. Round 1 is a full algorithmic coding assessment. Candidates who arrive without recent LeetCode practice drop here before any technical conversation begins.
• Networking theory without hands-on tracing. Describing TCP handshakes is not the same as having debugged a broken CNI plugin. Interviewers at AVP level detect the difference in the first follow-up question.
• Generic STAR stories with no numbers. "I improved team communication" with no named incident, no named tool, and no measurable result will not survive a follow-up. Every story needs at least one concrete figure.
• Interpreting the 10-day post-offer silence as rejection. JPMC is a large financial institution. Compensation approval and background verification run through multiple internal layers. Praveen waited 10 days and got the offer. Trust the process.
• Applying at AVP level without production incident ownership. The calibration question at this level is not "have you used Kubernetes?" but "what did you break, what did you fix, and how do you know it stayed fixed?" Interview stories without named incidents and personal accountability do not clear the AVP bar.

Real-World Data Points

• 5 rounds total in Praveen Singampalli's JPMC SRE AVP process
• Round 1 was a proctored HackerRank coding assessment, described in his words as "hell"
• Round 4 required every answer in strict STAR format, with live situational and real-time incident challenges added
• 10 days elapsed between the HR and offer discussion round and the formal offer letter arriving
• 140 reactions on the LinkedIn post, indicating high relevance across the DevOps and SRE hiring community
• Praveen's background: Ex-Verizon, Ex-Infosys, AWS Community Builder, 130K+ LinkedIn followers, 80K+ YouTube subscribers
• Interview cycle from the 2025 hiring season; breakdown published publicly ahead of 2026 hiring at JPMC

FAQ

How many rounds does a JPMorganChase SRE AVP interview have?

Based on Praveen Singampalli's public LinkedIn account, the process ran 5 rounds: a proctored HackerRank coding test, two technical interviews, a STAR-format behavioral round with real-time incident scenarios, and a final HR/offer discussion. Round count can vary by team and hiring manager.

Is DSA coding mandatory for JPMC SRE roles?

Yes. Praveen explicitly names Round 1 — a timed, proctored HackerRank assessment — as the hardest part of the process. For SRE and infrastructure roles at product companies in 2026, a working DSA foundation is mandatory alongside systems depth.

What format does JPMC use for behavioral interviews at AVP level?

JPMC uses strict STAR format (Situation, Task, Action, Result) with real-time situational challenges mixed in. Prepare specific, named stories with measurable outcomes — generic answers will not survive the follow-up probing that AVP-level interviewers apply.

How long does the JPMC offer letter take after Round 5?

In Praveen's case, 10 days after the HR and offer discussion. JPMC's internal approval and background verification process involves multiple layers; that window is standard, not a negative indicator.

What should I focus on for JPMC SRE technical rounds?

Praveen's framework points to three areas: Linux as a diagnostic instrument (/proc, signals, fd limits, log-based RCA), end-to-end network tracing including DNS and TCP failure analysis, and container internals (cgroups, namespaces, OOMKill, Kubernetes QoS classes). At AVP level, interviewers want incident depth and ownership, not tool familiarity alone.

Sources & credits

• https://www.linkedin.com/posts/praveen-singampalli_i-have-recently-cracked-jpmorganchase-sre-activity-7274437970273693696-5Omm